The long-term goal of this research program is to understand the mechanistic and evolutionary causes of variation in complex traits. The current focus is on mechanisms that appear to either suppress or promote variation. The primary experimental approach is to perform large-scale analyses of single-cell traits of the budding yeast, Saccharomyces cerevisiae. One line of work joins others in showing that cryptic genetic variation, kept suppressed until a perturbation reveals its phenotypic effects, is pervasive. This observation suggests that genetic interactions (epistasis) might be a major determinant of complex-trait variation. A second line of work joins others in suggesting that some clonal populations generate heterogeneity in order to hedge their bets against environmental uncertainty. The research program will follow these two lines of work. One set of projects aims to understand how epistasis contributes to natural variation in complex traits. Understanding the sources of variation in complex traits is a major goal in biomedical research because this knowledge impinges directly on the prospect of personalized medicine, for example the prediction of disease risk from an individual's genotype. If not taken into account, epistasis can confound such predictions. Epistasis is also important because it can constrain evolutionary adaptation to follow particular paths, making adaptation more predictable. This predictability could be valuable in the treatment of diseases that have a strong evolutionary component, such as microbial infections and cancer. Although epistasis has been well studied using lab- derived mutations, it has not been well studied in nature because most experimental designs have insufficient power to detect interacting loci. A key aim of this research program is to perform studies with dramatically increased power to detect interactions, for a large number of independent phenotypes, to gain a far richer view of the underlying causes of differences in complex traits. These studies will leverage recent progress in developing high-throughput, microscopy-based methods of quantifying many independent phenotypes, and they will create and use strains of S. cerevisiae that make searching for epistasis much more powerful. The other set of projects aims to understand the molecular mechanisms underlying a newly discovered bet-hedging phenomenon whereby clonal populations of S. cerevisiae contain fast-growing cells that are sensitive to acute stress and slow-growing cells that are tolerant of acute stress. Molecular mechanisms of this kind of adaptive heterogeneity are poorly understood, especially in eukaryotes, so the opportunity to study such a system in a model eukaryote with powerful genetic, molecular and cell-biological tools could lead to major advances. A candidate pathway for controlling the heterogeneity in growth and stress resistance will be studied. In addition, natural variation in growth-rate distributions between S. cerevisiae strains will be mapped, in an effort to understand how ecological pressures shape bet-hedging mechanisms. The two lines of work converge here because epistatic interactions appear to dominate the genetic basis of differences in growth-rate variance.